Savella in Treatment for Provoked Vestibulodynia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Savella in Treatment for Provoked Vestibulodynia (NCT01304589), a na phase 3 study. The primary outcome is Pain Rating Index.

2. Clinical Question

3. Trial Design

The registry describes an open-label, single-group trial of milnacipran in women with provoked vestibulodynia. It lists the phase as phase 3, while the intervention description calls the study an exploratory trial. Allocation is not applicable and masking is none. These design features mean that participants and investigators were not blinded to treatment, and there was no randomized control group against which to compare the treated group.

The planned study duration is described as 18 weeks, including screening and baseline assessment, dose escalation, and a stable-dose phase. The intervention description says eligible patients were treated with 200 mg/day, or the maximum tolerated dose, for a total of 12 weeks. The registry lists enrollment as 31. For the primary outcome, the analysis population is described as all 22 eligible subjects who received study medication. Reasons why the enrollment and primary-outcome populations differ are not reported in the registry record.

The primary endpoint was the Pain Rating Index measured at baseline and at 18 weeks. Secondary outcomes listed in the record are tampon pain, coital pain, and 24-hour vulvar pain. Although these repeated measurements can describe how scores changed within the treated group, a before-and-after comparison alone cannot distinguish a treatment-associated change from changes that might also occur over time for other reasons. The record does not report a control condition or a randomized treatment contrast.

The registry lists the study as completed, with a start date of 2010-10 and primary completion date of 2012-09. Those dates describe the registered study timeline; they do not provide outcome estimates. The posted registry information does not report the detailed statistical analysis plan, the rationale for the planned enrollment, or whether there were interim analyses.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: 22 subjects were eligible and received study medication. Analysis population included all 22 eligible subjects.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Pain Rating Index, measured over 18 weeks.

The primary endpoint is the Pain Rating Index, a component of the McGill Pain Questionnaire intended to capture sensory and affective components of pain. The registry describes its range from 0, meaning no pain, to 45, meaning severe pain. Scores were used to measure mean values at baseline and at 18 weeks after treatment.

A pain measure is relevant to the stated study question because the trial aimed to assess pain in women with provoked vestibulodynia. The index combines sensory and affective dimensions, so it represents more than a description of one isolated pain quality. Its interpretation nonetheless depends on the score’s actual distribution, the amount and direction of change, and the precision of the estimate. The registry record does not report the baseline or 18-week means, standard deviations, a mean change, or an uncertainty interval for this endpoint.

Because this was a single-group study, a comparison of baseline and follow-up scores would describe change over time among treated participants. It would not, by itself, establish what would have happened over the same period without treatment. The record identifies the measurement times but does not report the specific statistical procedure used to compare them or whether the endpoint was analyzed as follow-up values, change scores, or another summary.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation begins with identifying what comparison the design can support. This trial is described as single-group, open-label, and without masking, with no randomized comparator. Its primary measurement was taken at baseline and at 18 weeks. Those features support a description of scores at two time points among participants in the treated group, but do not create a randomized estimate of the difference treatment itself caused.

What does a mean describe?

The primary endpoint is a continuous measure, and the registry identifies the reported summary as a mean with standard deviation. A mean describes the average score in the analyzed group. A standard deviation describes how spread out individual scores are around that average. They answer different questions: the mean locates the group average, while the standard deviation describes variation among participants. Neither alone states how precisely the group mean has been estimated; precision also depends on the analysis population and the statistical method.

The Pain Rating Index has a stated range from 0, no pain, to 45, severe pain. A mean difference can be easier to interpret when its scale and direction are clear, but the registry record does not report the numerical means or a change estimate. Nor does it report whether the analysis focused on separate time-point means or on within-person changes. The mere presence of measurements at baseline and follow-up does not identify which analysis was used.

Why does paired data matter?

General statistical principle: when the same participants are measured at two times, each person’s follow-up value is naturally linked to that person’s baseline value. An analysis of within-person change can use this pairing. It can differ from treating the two sets of measurements as unrelated groups because the paired analysis accounts for the relationship between an individual’s repeated scores. The registry record does not report whether the primary analysis used a paired method, how change was defined, or how missing follow-up values were handled.

A confidence interval, when reported, gives a range of values compatible with the data and the specified statistical model. Its width conveys precision: a narrower interval indicates a more precise estimate under that model, while a wider interval indicates greater uncertainty. A p-value addresses compatibility with a specified null hypothesis under model assumptions; it does not measure the size or importance of an effect. The registry record does not report a confidence interval or p-value for the primary endpoint, so neither precision nor evidence against a null hypothesis can be assessed from this record.

Why is a single-group change not a controlled effect?

General statistical principle: a randomized control group helps estimate what might have occurred over the same period in the absence of the studied intervention. In a single-group before-and-after design, the observed change combines any treatment-associated change with other time-related influences and measurement variation. Open-label measurement can also be affected by awareness of treatment, depending on the outcome and assessment process. These are limitations of what the design can identify, not claims that a particular source of bias occurred in this study.

There is also a distinction between a statistical association and causation. If an outcome changed after treatment, timing alone would not demonstrate that treatment produced the change. A controlled comparison can better address this question, especially when treatment assignment is randomized and assessment procedures are specified. The registry record does not report a separate control group for this trial.

What do sample-size illustrations mean?

Any sample-size figures associated with this record are planning illustrations, not observed power. The stated illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. The record specifies that such figures are participants per arm for non-time-to-event outcomes. Those assumptions do not match the registered single-group design, which has no two-arm randomized comparison. Therefore the illustrations should not be read as the study’s enrollment target, as a description of its actual design, or as proof that the completed study had a particular probability of detecting an effect. The trial’s own sample-size justification is not reported in the registry record.

General statistical principle: sample-size calculations depend on the target effect, outcome variability, significance threshold, allocation, and analysis method. Changing those assumptions changes the planning result. A retrospective calculation based on an observed effect is especially sensitive to the observed estimate and does not replace a prospectively justified design. The record does not report the numerical assumptions or planning results needed to evaluate a particular illustration in detail.

Why can over-dispersion matter?

General statistical principle: over-dispersion means that observed variability is greater than a model assumes. If a model understates variability, uncertainty can be understated and tests or intervals may be too optimistic. For a continuous mean endpoint, the standard deviation is one way to describe observed spread; other outcome types may require different variance assumptions. The registry record does not report an over-dispersion assessment, and the primary endpoint is described as a continuous mean. This general issue does not establish that over-dispersion occurred in this trial.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Milnacipram")
n   <- c(22)
m   <- c(12.3)
sp  <- c(13.3)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Milnacipram  22  12.300  13.300  6.403  18.197

11. Multiplicity, Missing Data and Analysis Populations

The registry lists the Pain Rating Index as the primary outcome and names tampon pain, coital pain, and 24-hour vulvar pain as secondary outcomes. It also records four outcome measures and four analyses. Examining multiple outcomes can increase the chance of finding at least one apparently unusual result even when no true effect exists. General statistical principle: multiplicity procedures can control or otherwise address this concern, but the appropriate approach depends on which outcomes are considered confirmatory and how the analyses are organized. The registry record does not report whether multiplicity was addressed, what testing hierarchy was used, or how the four analyses correspond to the four measures.

The primary-outcome population is reported as all 22 eligible subjects who received study medication. The registry enrollment figure is 31, and the adverse-event count uses 31 at risk. These are distinct recorded counts. Reasons for the difference, participant flow, and the disposition of participants without primary endpoint data are not reported in the registry record.

Missing data can affect both the estimate and its uncertainty, particularly when participants with missing follow-up measurements differ from those with observed measurements. General statistical principle: the impact depends on the amount, pattern, and likely reasons for missingness, as well as the analysis method. The registry record does not report the number of missing primary measurements, the method used to handle missing data, or sensitivity analyses assessing how missingness could affect the result.

The record identifies an analysis population but does not report whether all enrolled participants, all treated participants, or another defined population contributed to each secondary analysis. It also does not report a detailed statistical analysis plan. These omissions limit assessment of whether outcome definitions and analytic choices were specified before results were examined.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event record lists 31 participants at risk for serious adverse events and 0 affected. This is a count of participants recorded as experiencing a serious adverse event, not a measure of the probability that such an event could occur in other settings. The record lists no serious adverse-event terms. The deaths value is not reported in the registry record.

The same safety record lists 31 participants at risk for other adverse events and 15 affected. These figures describe the registry’s recorded counts; they do not identify the particular events, their severity, duration, or relationship to treatment. The specific other adverse-event terms are not reported in the registry record. The registry does not provide enough detail to determine whether events were counted once per participant or how repeated events were handled.

The safety denominator of 31 differs from the 22 eligible subjects described in the primary-outcome analysis population. The reason for this difference is not reported in the registry record, so the safety and primary-outcome populations should not be assumed to be identical. Counts alone also do not establish whether an event was caused by the intervention. General principle: adverse-event summaries describe what was recorded during observation, while attributing an event to treatment requires additional clinical and comparative assessment.

13. What the Result Does, and Does Not, Mean

The registry record establishes that an open-label, single-group study evaluated milnacipran in women with provoked vestibulodynia and identified the Pain Rating Index at baseline and 18 weeks as its primary measure. It also describes the study’s flexible-dose regimen, secondary pain outcomes, and the population listed for the primary analysis. These details clarify the question the study was designed to address and the framework in which its measurements were collected.

The numerical primary endpoint results are not reported in the registry record. There is therefore no reported estimate in this record from which to describe the direction or size of change in the Pain Rating Index, and no reported interval from which to judge precision. The registry information cannot establish whether mean pain scores changed, how consistent any change was across participants, or whether an analysis met a statistical threshold.

Even if a within-group change estimate were available, the single-group design would limit causal interpretation. A before-and-after difference does not provide the same evidence as a randomized comparison with a control group. It cannot isolate the intervention’s effect from other changes over time. Open-label assessment and the difference between enrolled participants and the primary analysis population add further considerations, though the record does not state why those counts differ.

The safety counts indicate that the record lists no participants affected by serious adverse events among 31 at risk and 15 affected by other adverse events among 31 at risk. The nature of the other events and the death count are not reported in the registry record. Those counts should be read as recorded safety information, not as proof about the cause of events or a complete characterization of safety.

Overall, the record supports a description of the design, outcomes, and recorded safety counts, but does not provide the quantitative primary results needed to judge the observed pain trajectory or its precision. It also does not provide a randomized contrast. These are limits on what can be concluded from the registered information, rather than evidence for or against a treatment effect.

14. Why This Trial Matters Statistically

This trial illustrates why the design behind an outcome estimate matters as much as the outcome name. A baseline and follow-up pain score can be clinically relevant and straightforward to describe, but the statistical question changes when the study has no randomized control group. In a single-group design, the observed course of scores cannot be separated from other influences over time by comparison with a concurrent control.

It also highlights the importance of keeping analysis populations distinct. The registry reports enrollment of 31, a primary-outcome population of 22 eligible subjects who received study medication, and safety counts based on 31 at risk. Those counts refer to different recorded contexts, and the reasons for their difference are not reported. Readers need denominators that correspond to the outcome being interpreted; otherwise, a count can be attached to the wrong population.

The endpoint’s scale and summary are explicitly described: the Pain Rating Index spans 0 to 45, and mean values were measured at baseline and 18 weeks. Yet a measurement definition is not a result. Without numerical outcome estimates and uncertainty, the registered description cannot show the magnitude or precision of change. This distinction is central to statistical reporting: design and endpoint details explain what was intended to be measured, while estimates and intervals are needed to quantify what the observed data indicate.

The record also demonstrates why planning calculations require context. The sample-size illustrations use a two-arm framework, whereas this registered study is single-group. Such illustrations are not observed power and do not retrospectively validate the study’s size. A calculation only answers the planning question encoded by its assumptions; it cannot create a comparator or repair a mismatch between assumptions and design.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The principal limitation for interpretation is that the registry record does not report numerical results for the primary Pain Rating Index endpoint. Baseline and follow-up means, standard deviations, change estimates, confidence intervals, and p-values are not reported in the registry record. The primary question therefore cannot be answered quantitatively from the registered information.

The single-group, open-label design has no randomized control arm. A within-group change over time, if reported, would not by itself identify the change attributable to milnacipran. The record does not report a control condition, masking, or a randomized comparison. These design features constrain causal interpretation regardless of the endpoint’s measurement quality.

The registry lists 31 enrolled participants, while the primary-outcome population comprises 22 eligible subjects who received study medication; safety counts use 31 at risk. Reasons for the different populations, the participant flow, and the handling of missing outcome measurements are not reported in the registry record. The relationship between eligibility, treatment exposure, and inclusion in each analysis cannot be fully assessed.

Although several secondary pain outcomes are named, their results and the relationship between the four recorded analyses and four outcome measures are not reported in the registry record. Multiplicity procedures, the full analysis plan, the sample-size justification, interim analyses, and detailed missing-data methods are also not reported in the registry record. These omissions limit assessment of the prespecification, robustness, and completeness of the reported analyses.

The safety information is limited to counts of participants at risk and affected. The registry does not report the specific other adverse-event terms or a death count. These limitations mean the record gives only a partial statistical and safety account, rather than a complete basis for estimating outcome effects or evaluating the full range of observed events.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Savella in Treatment for Provoked Vestibulodynia — NCT01304589.
• Registry record last updated 2018-06-11; status last verified 2018-05.
• Results first posted 2014-04-02.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry record describes a completed phase 3, open-label, single-group exploratory trial of milnacipran for women with provoked vestibulodynia. The intervention description specifies flexible dosing, with treatment at 200 mg/day or the maximum tolerated dose for 12 weeks within an 18-week study design. Enrollment is listed as 31.

The primary outcome was the Pain Rating Index, measured at baseline and 18 weeks. Its analysis population is described as all 22 eligible subjects who received study medication. Tampon pain, coital pain, and 24-hour vulvar pain are listed as secondary outcomes. Numerical primary and secondary outcome results are not reported in the registry record.

The adverse-event record lists 0 affected by serious adverse events among 31 at risk and 15 affected by other adverse events among 31 at risk. The specific other events and death count are not reported in the registry record. The registry does not report the reasons for the differing primary-outcome and safety populations.